24 February: DCRLS -- Feb 24, 4pm: Nicholas Asher
Dublin Computational Linguistics Research Seminar: Index of February 2006 | Dublin Computational Linguistics Research Seminar - Index of year: 2006 | Full index
The next Computational Linguistics Research Seminar will feature a
talk by Prof. Nicholas Asher, of University of Texas at Austin,
Department of Philosophy
The seminar will take place
- on Friday February 24,
- at 4pm
- in Room B1.09
- in UCD's School of Computer Science and Informatics, Belfield Campus,
( on the map at http://www.ucd.ie/maps/campusmap05.jpg we are
building "22" an inch
to the left of the topleft corner of the central lake)
All are welcome to attend.
-------
Prof. Nicholas Asher
Title:
Context and Complex Types in Predication
Abstract:
For some time now, lexical semantics has made free use of types to in
an account of predication and to analyze coercion effects and
copredication of the sort in (1) and (2) (e.g., Pustejovsky 1995).
(1) Pat enjoyed the cigar.
(2) Lunch was delicious but took forever.
And while typed feature structure formalisms (TFSs)(of the sort
developed in Carpenter (1991)) are commonly used in linguistics, they
do not offer ready-made solutions to coercion One reason for this is
the need to introduce complex, non-functional types of the sort
countenanced in Asher and Pustejovsky (2001) or Asher and Lascarides
(2001), whereas in a standard modal semantics for TFSs (Blackburn
1992), types are treated as atoms. The models of complex types needed
for copredication are not well understood, and I will talk about a
model proposed by Cooper and also one proposed by Pustejovsky,
A second reason for an unhappiness with current theories of typed
predications is that many broadly speaking predicative constructions
like the relative predication in (3) (Asher 2006), the genitive
construction in (4) (Asher and Denis 2004) or certain uses of aspectual
verbs in (5) (Danlos 2005)
resist any analysis in which a static set of types is used to guide
the predication.
(3) This book as a paddle is useless.
(4)[Context: Picasso and Raphael both painted a mother with a
child.]
Picasso's mother is bright and splashy ---a typical cubist
rendition. It makes Raphael's mother look so somber.
(5a) ?Yesterday, Sabrina began with the kitchen. She then proceeded to
the living room and bedroom and finished up with the bathroom.
(5b) Yesterday Sheila cleaned her house. She began with the kitchen.
She then proceeded to the living room and bedroom and finished up with
the bathroom.
(5c) Last week Julie painted her house. She began with the kitchen.
She then proceeded to the living room and bedroom and finished up with
the bathroom.
I argue that we need a {\em contextualist model of type driven lexical
semantics} to handle such examples with a particular analysis of
complex types to handle these phenomena.
-------
The support for DCLRS2005-6 given by the School of Computer Science and
Informatics, UCD Dublin,
is gratefully acknowledged.
_______________________________________________
cogsci mailing list
cogsci@cs.tcd.ie
https://lists.cs.tcd.ie/mailman/listinfo/cogsci
Dublin Computational Linguistics Research Seminar - Index of February 2006 | Index of year: 2006 | Full index